The Routing Algorithm Study Of LEO Satellite Network

Routing is always a key technology in communications network,good or bad routing arithmetic will influence the performanceof communications network and the quality of communications.Although there is a lot of mature routing arithmetics in earthnetwork, which can not be used in satellite networks for thecharacteristic of the satellite communications. Some routingarithmetic that is fit for the satellite communications must bedesigned.The number of satellite and the topology configuration aredifferent in satellite network, so the link fashion is different.When the number of satellite in satellite net work is small,Satellite can be linked with each other by EGS(earth-gate-station). The connect way is logical for the highreliability, low-cost and the simple technique. The way linkedby EGS need a large mount of EGS and high-cost and some specialEGS to be located in ocean, desert and enemy countries, but thatis impossible. So it is not realizable to connect satellite witheach other in the satellite network only by earth-station. Theway satellite network linked by ISL (Inter-Satellite-Links) canmake up the weakness of EGS. These must be different routingalgorithms for different linked manner to seek the bestcommunication route.For LEO satellite network who can provide communicationservice to the global users with no inter—satellite links butmake use of the high--speed and low—cost ground-backbonenetwork to form a "siphon-manner" communication network, TheLEO satellite data communication system is one of the typicaldeputy. Real time communication and S&F(Store/Forward)communications exist at the same time in this system, whichneeds ground network and ground gateway's help forwell-balanced run. In fact, nations can not built gateways inglobal regions. These gateways must be restricted in their owncountry. According to the station facts and the characteristicof S&F(Store/Forward) communications, in This chapterdissertation, we propose the shortest delay routing algorithmwhich is fit for the LEO satellite data communication system.When During the communication links are building, thealgorithm can judge the communication is real time or S&F;Ifit is S&F, the algorithm can work out the shortest time delayand find out the optimal route。Using the algorithm can fullyutilize the valued satellite resource and the better serviceperformance can be provided for user. The computer simulationresults indicate the validity of this algorithm.The routing algorithm of the satellite communicationssystem with ISL is multiplicate, particular introduction ofthese algorithm is given in chapter 3 and section 2.For thecharacteristic of satellite communications, How to solve thedelay-constrained routing problem in designing algorithm mustbe considered, which has not been considered in many routingalgorithm that only aim at the short route. Only distributedarithmetic is adopted in DQR to solve the delay-constrainedrouting in LEO satellite network. However, the small LTIparameter and account inaccurate exist in routing selected bythe DQR algorithm. The shorter LTI parameter leads to worsevalidity of the routing, that is, the selected routing abatequickly again;account inaccurate leads to worse veracity of therouting, that is, the selected routing can not satisfy thedelay-constrained request probably.Directing at the weaknesses of the DQR arithmetic, LCODCR(LTI Cost Optimization Delay Constrained Routing) is proposedto resolve the delay-constrained routing algorithm in LEOsatellite network. Adopting the centralized algorithm, LCODCRcan obviously reduce the route setup time. Shorter setup-timeof route means the longer effective life-time of route. Theaccurate result can be acquired in this arithmetic, so that theveracity and the validity of the delay-constrained routingalgorithm can be enhanced, however the complicacy of thearithmetic also is increased, which leads to the calculationburden of the LEO satellite node and the question of the routingcalculation invalidation in LEO satellite network.The countermeasure is to ensure the veracity and validityof the LCODCR arithmetic fully utilizing the periodicity,anticipation and fixity of the LEO satellite network to reducethe complicacy of the LCODCR arithmetic. The main method isadopting various simplification arithmetic to reduce thecomplexity of the LCODCR arithmetic. The ameliorated strategyin which the dynamic approximate method and the compressedmethod are used in satellite network is put forward;in thedynamic approximate method, the dynamic approximate factorestimate strategy is put forward, which is used to reduce theveracity of the arithmetic so the complexity of the arithmeticcan be reduced;in the compressed method, the compressed factoris put forward, which is used to reduce the success of thearithmetic so the complexity of the arithmetic can be reduced.In one word, the purpose is to reduce the amount of the routinganswer to reduce the complexity of the arithmetic-selecting thebest answer. The best answer has the same probability with theaccurate answer and can be proved in theory as well as thefeasibility of the method. The simulate model is built up tocompare the LCODCR arithmetic with the DQR arithmetic. Theresult proves that the LCODCR arithmetic can resolve the delayconstrained routing problem better in satellite network.